Deep Learning in Diagnosis of Dental Anomalies and Diseases: A Systematic Review

Diagnostics, Journal Year: 2023, Volume and Issue: 13(15), P. 2512 - 2512

Published: July 27, 2023

Deep learning and diagnostic applications in oral dental health have received significant attention recently. In this review, studies applying deep to diagnose anomalies diseases image material were systematically compiled, their datasets, methodologies, test processes, explainable artificial intelligence methods, findings analyzed. Tests results involving human-artificial comparisons are discussed detail draw the clinical importance of learning. addition, review critically evaluates literature guide further develop future field. An extensive search was conducted for 2019–May 2023 range using Medline (PubMed) Google Scholar databases identify eligible articles, 101 shortlisted, including diagnosing (n = 22) 79) classification, object detection, segmentation tasks. According results, most commonly used task type classification 51), panoramic radiographs 55), frequently performance metric sensitivity/recall/true positive rate 87) accuracy 69). Dataset sizes ranged from 60 12,179 images. Although algorithms as individual or at least individualized architectures, standardized architectures such pre-trained CNNs, Faster R-CNN, YOLO, U-Net been studies. Few AI method applied tests comparing human 21). is promising better diagnosis treatment planning dentistry based on high-performance reported by For all that, safety should be demonstrated a more reproducible comparable methodology, with information about applicability, defining standard set metrics.

Language: Английский

---

Deep learning for classifying the stages of periodontitis on dental images: a systematic review and meta-analysis

BMC Oral Health, Journal Year: 2023, Volume and Issue: 23(1)

Published: Dec. 19, 2023

Abstract Background The development of deep learning (DL) algorithms for use in dentistry is an emerging trend. Periodontitis one the most prevalent oral diseases, which has a notable impact on life quality patients. Therefore, it crucial to classify periodontitis accurately and efficiently. This systematic review aimed identify application DL classification assess accuracy this approach. Methods A literature search up November 2023 was implemented through EMBASE, PubMed, Web Science, Scopus, Google Scholar databases. Inclusion exclusion criteria were used screen eligible studies, studies evaluated by Grading Recommendations Assessment, Development Evaluation (GRADE) methodology with QUADAS-2 (Quality Assessment Diagnostic Accuracy Studies) tool. Random-effects inverse-variance model perform meta-analysis diagnostic test, pooled sensitivity, specificity, positive likelihood ratio (LR), negative LR, odds (DOR) calculated, summary receiver operating characteristic (SROC) plot constructed. Results Thirteen included meta-analysis. After excluding outlier, LR DOR 0.88 ( 95%CI 0.82–0.92), 0.82 0.72–0.89), 4.9 3.2–7.5), 0.15 0.10–0.22) 33 19–59), respectively. area under SROC 0.92 0.89–0.94). Conclusions DL-based high, approach could be employed future reduce workload dental professionals enhance consistency classification.

Language: Английский

---

Caries-segnet: multi-scale cascaded hybrid spatial channel attention encoder-decoder for semantic segmentation of dental caries

Biomedical Engineering / Biomedizinische Technik, Journal Year: 2025, Volume and Issue: unknown

Published: March 6, 2025

Dental caries is a prevalent oral health issue around the world that leads to tooth aches, root canal infections, and even extractions. Existing dental diagnosis models may misdiagnose disorder take more time segment caries. This research work aims provide an in-depth analysis of spatial channel attention mechanism techniques used for semantic segmentation in encoder-decoder network. For effective performance, implements novel accurately. Deep Fully Connected Residual Block (DFCR) designed relevant features without loss significant information. A Hybrid Spatial Channel Attention (HSCA) module developed combining with help multi-scale cross-dimensional features. The proposed methodology performs better than other cutting-edge algorithms by achieving 96.63 % accuracy, 95.77 dice score, 96.28 Intersection over Union (IOU) score dataset, 96.93 95.21 value, 96.1 IOU Tufts dataset. model facilitates detection cavities precisely at earlier stage images. provides accurate assisting medical professionals.

Language: Английский

---

Breast Cancer Segmentation in Mammograms using Antlion Optimization and CNN/GRU Architectures

2022 International Wireless Communications and Mobile Computing (IWCMC), Journal Year: 2024, Volume and Issue: unknown, P. 1030 - 1035

Published: May 27, 2024

Language: Английский

---

Deep-learning based fusion of spatial relationship classification between mandibular third molar and inferior alveolar nerve using panoramic radiograph images

Biomedical Signal Processing and Control, Journal Year: 2024, Volume and Issue: 100, P. 107059 - 107059

Published: Oct. 30, 2024

Language: Английский

---

Primary Methods and Algorithms in Artificial-Intelligence-Based Dental Image Analysis: A Systematic Review

Algorithms, Journal Year: 2024, Volume and Issue: 17(12), P. 567 - 567

Published: Dec. 11, 2024

Artificial intelligence (AI) has garnered significant attention in recent years for its potential to revolutionize healthcare, including dentistry. However, despite the growing body of literature on AI-based dental image analysis, challenges such as integration AI into clinical workflows, variability dataset quality, and lack standardized evaluation metrics remain largely underexplored. This systematic review aims address these gaps by assessing extent which technologies have been integrated specialties, with a specific focus their applications imaging. A comprehensive was conducted, selecting relevant studies through electronic searches from Scopus, Google Scholar, PubMed databases, covering publications 2018 2023. total 52 articles were systematically analyzed evaluate diverse approaches machine learning (ML) deep (DL) reveals that become increasingly prevalent, researchers predominantly employing convolutional neural networks (CNNs) detection diagnosis tasks. Pretrained demonstrate strong performance many scenarios, while ML techniques shown utility estimation classification. Key identified include need larger, annotated datasets translation research outcomes practice. The findings underscore AI’s significantly advance diagnostic support, particularly non-specialist dentists, improving patient care efficiency. AI-driven software can enhance accuracy, facilitate data sharing, support collaboration among professionals. Future developments are anticipated enable patient-specific optimization restoration designs implant placements, leveraging personalized history, tissue type, bone thickness achieve better outcomes.

Language: Английский

---

Autonomous dental treatment planning on panoramic x-ray using deep learning based object detection algorithm

Multimedia Tools and Applications, Journal Year: 2023, Volume and Issue: 83(14), P. 42999 - 43033

Published: Oct. 13, 2023

Language: Английский

---

A cropping algorithm for automatically extracting regions of ınterest from panoramic radiographs based on maxilla and mandible parts

International Journal of Information Technology, Journal Year: 2023, Volume and Issue: 15(7), P. 3631 - 3641

Published: Aug. 22, 2023

Language: Английский

---

DA-Net: A classification-guided network for dental anomaly detection from dental and maxillofacial images

Journal of King Saud University - Computer and Information Sciences, Journal Year: 2024, Volume and Issue: 36(9), P. 102229 - 102229

Published: Oct. 31, 2024

Language: Английский

---

Classification of Impacted Teeth from Panoramic Radiography Using Deep Learning

Lecture notes in computer science, Journal Year: 2024, Volume and Issue: unknown, P. 257 - 270

Published: Dec. 31, 2024

Language: Английский

---